Brave Search API

Search Brave’s public index with your ScrapeBadger key. Get organic results and source-provided rich blocks, including knowledge blocks when available, or use the image, video and news endpoints. You do not need a Brave account or a separate Brave API key.

Endpoints and credits

Endpoint	Credits per nonempty response
Web search	7
Images	7
Videos	7
News	7
Autocomplete	2
Locale choices	0

Confirmed empty results and failed requests are unbilled. Read X-Credits-Used for the charge on a response. Keys restricted to particular APIs need the brave scope.

Quickstart

curl -G "https://scrapebadger.com/v1/brave/search" \
  -H "X-API-Key: YOUR_API_KEY" \
  --data-urlencode "query=coffee machine" \
  --data-urlencode "country=us" \
  --data-urlencode "freshness=week"

import os
import httpx

response = httpx.get(
    "https://scrapebadger.com/v1/brave/search",
    headers={"X-API-Key": os.environ["SCRAPEBADGER_API_KEY"]},
    params={"query": "coffee machine", "page": 1},
    timeout=120,
)
response.raise_for_status()
data = response.json()
for result in data["results"]:
    print(result["position"], result["title"], result["url"])

const response = await fetch("https://scrapebadger.com/v1/brave/search?" +
  new URLSearchParams({query: "coffee machine", page: "1"}), {
  headers: {"X-API-Key": process.env.SCRAPEBADGER_API_KEY},
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const data = await response.json();
console.log(data.results);

Pagination and filters

Web and video search accept a 1-based page ordinal (page=1 through 10). Use has_next and next_page to continue. count trims one native page; it does not change Brave’s page boundaries. You may receive fewer results than requested. Images and news return their initial source batch and have no page parameter. All search endpoints accept country, ui_language, and safe_search=off|moderate|strict. Use the free locale reference for valid codes. UI language controls interface labels and date formatting; it is not a result-language filter. country=all selects all regions. Web search also accepts freshness=any|day|week|month|year and spellcheck. Read query_info for source spelling corrections and query relaxation. Brave can broaden a query even with spellcheck disabled. Unsupported parameters return 422.

Result fields

Every search response includes results, result_count, query_info, source_url, country, ui_language, safe_search, and the observation time as both fetched_at (ISO UTC) and fetched_utc (Unix seconds). Results include title, URL, position, snippets and source metadata. Web search also returns FAQ, discussions, related searches and inline media when Brave supplies them. Rich modules vary by query; an empty module array means Brave did not supply that module. Original image URLs, dimensions, video metadata and publisher profiles remain in their source-specific result fields. Relative article ages remain display strings. Absolute publication timestamps appear only when the source supplies a defensible date; observation time is separate from publication time. Brave AI Mode and generated answers are outside this API.

CLI and MCP

scrapebadger brave search "coffee machine" --freshness week --page 2
scrapebadger brave images "golden retriever" --count 10
scrapebadger brave videos "espresso tutorial" --page 2
scrapebadger brave news "technology" --count 10
scrapebadger brave locales

MCP tools: brave_search, brave_images, brave_videos, brave_news, brave_autocomplete, brave_locales. ScrapeBadger is independent and is not affiliated with or endorsed by Brave.

Overview

API reference

Endpoints and credits

Quickstart

Result fields

CLI and MCP

​Endpoints and credits

​Quickstart

​Pagination and filters

​Result fields

​CLI and MCP

Endpoints and credits

Quickstart

Pagination and filters

Result fields

CLI and MCP